Hacker News Reader

Comment by Jun8 | Hacker News Reader

The entry in Stanford Encyclopedia of Philosophy has a good summary of the Language of Thought Hypothesis. https://plato.stanford.edu/entries/language-thought/

Jerry Fodor, who advocated an internal mental language, was seen as an old crank when I studied Linguistics in late 90s; I just looked him up and, god, he died recently in 2017. Here’s a New Yorker article from that time: https://www.newyorker.com/culture/postscript/jerry-fodors-en....

View on HN

Complaint not targeted at you personally but - why does thought have to take place in language at all? I sometimes reason using physical or kinesthetic metaphors that would be impossible to capture perfectly in language.

and why does thought have to have one form? I think in both language and visual representations and i switch between them depending on the task and i even pipe the input from one representation into the other representation.

This is related but not really the same thing. The Language of Thought may or may not be closely related to human language. (Some animals may also have concepts and compositional mental representations.)

I didn't read the SEP article, but is the language of thought hypothesis perhaps already verified by our ability to do some sort of logic on neural network embeddings? E.g. queen-royal=woman.

Fundamentally LoT is a hypothesis about human (and possibly animal) cognition, not a hypothesis about what artificial neural network models can or can't do. But you can see Fodor and Pylyshyn's thoughts on 'connectionist' models (as they were known at the time) in this extensive article from 1988: https://ruccs.rutgers.edu/images/personal-zenon-pylyshyn/pro... The discussion on p. 14 is particularly relevant.

Reply on news.ycombinator.com